[Analytical quality of assays and comparison of procedures for the sweat test].

Sweat test measuring the chloride ion (Cl(-)) concentration in sweat is a tool for the cystic fibrosis (CF) diagnosis. We evaluated analytical criteria of different available methods and compared them into five hospitals and throught a national quality control program. Sweat tests were performed by stimulation using pilocarpine iontophoresis, sweat collection and measurement of sweat Cl(-) (mmol/L) by titration (colorimetric or coulometric end-point) or by in situ direct potentiometry using a chloride-selective electrode. Indirect determination by sweat conductivity measurement was expressed in mmol/L sodium chloride (NaCl) equivalents (Eq). Linearity range was demonstrated for all measurement procedures in the range 10 to 120 mmol/L. Intra-laboratory coefficients of variation (CVs) were <5% for values between 10 and 100 mmol/L. Inter-laboratory CVs were <3% only for conductivity measurement whatever the range. The comparison of results obtained for a same sweat sample, simultaneously by coulometric and conductivity measurements, demonstrated a first degree linear distribution between 30 to 60 mmol/L Cl(-) allowing us to establish an analytical correspondence table for this range. Thus, calculated values for 30, 40 and 60 mmol/L Cl(-) were respectively 57, 66 and 84 mmol/L NaCl Eq. In conclusion, comparison of methods highlighted that the less the sweat test is automatically controlled, the more the operator influence on results quality is important. Our study supports that sweat test result <50 mmol/L NaCl Eq is unlikely with CF diagnosis in absence of clinical arguments.